| noonedeadpunk | bjoernt: no, eom tags are static ones and created on stable/<brach> before it's deleted and replaced with unmaintained/<branch> | 07:36 |
|---|---|---|
| noonedeadpunk | and the timeout in https://zuul.opendev.org/t/openstack/build/f2c7abed4efe482fb8bbb403dce6cdb8/log/job-output.txt#6190 has nothing to do with deb822 at least | 07:37 |
| jrosser | I found yesterday that even wget/curl against the rabbitmq repo would just hang | 07:40 |
| jrosser | so yes it is unrelated | 07:40 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_octavia stable/2024.1: Return `amphora` provider back https://review.opendev.org/c/openstack/openstack-ansible-os_octavia/+/928999 | 11:01 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron stable/2024.1: Use cgroupsv2 for L3 cleanup https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/929001 | 11:04 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron stable/2023.2: Use cgroupsv2 for L3 cleanup https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/929002 | 11:05 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_neutron stable/2023.1: Use cgroupsv2 for L3 cleanup https://review.opendev.org/c/openstack/openstack-ansible-os_neutron/+/929003 | 11:05 |
| frickler | noonedeadpunk: can you do a quick check on jrosser's changes on the cycle highlights? https://review.opendev.org/c/openstack/releases/+/928884 | 14:04 |
| noonedeadpunk | sure, done | 14:18 |
| frickler | thx | 14:36 |
| spatel | noonedeadpunk morning! | 14:37 |
| noonedeadpunk | o/ | 14:37 |
| spatel | Sorry yesterday I left IRC so not able to finish conversation. | 14:40 |
| spatel | My Question was what issues you encounter during OVN test or POC? | 14:40 |
| spatel | Because I am planning to deploy OVN for 500 compute nodes and wanted to make sure I won't hit any unknown issues.. | 14:40 |
| spatel | so far it looks good in my POC lab testing | 14:41 |
| noonedeadpunk | ah, so we had quite a complex usecase, with nested networking, where at some level we mark geneve network as public, then create a router, and another private network | 14:42 |
| noonedeadpunk | and then FIP does not work for this geneve external net to layered private network | 14:42 |
| noonedeadpunk | consider corner case | 14:42 |
| spatel | But what was the motive behind this kind of case? | 14:43 |
| noonedeadpunk | um, we wanted some software firewall in between | 14:47 |
| noonedeadpunk | so that even firewall won't have access to public networks, so it should be on private one behind the router | 14:48 |
| spatel | hmm | 14:58 |
| opendevreview | Dmitriy Rabotyagov proposed openstack/openstack-ansible-os_horizon master: Fix custom policies rollout https://review.opendev.org/c/openstack/openstack-ansible-os_horizon/+/929125 | 15:01 |
| noonedeadpunk | spatel: and yes, we can directly attach ports from this geneve "external" net, which is a valid workaround | 16:25 |
| noonedeadpunk | but FIPs do not work :D | 16:25 |
| noonedeadpunk | and they did in OVS | 16:25 |
| spatel | hmm | 16:25 |
| spatel | I don't have that kind of requirement so assuming it will works for me.. | 16:26 |
| noonedeadpunk | https://bugs.launchpad.net/neutron/+bug/2077879 | 16:26 |
| noonedeadpunk | also - if you check their bugs - there will be some nits about fips/routers overall when it's not plain straightforward setup | 16:46 |
| noonedeadpunk | and master is already quite far from 2024.1 in terms of the driver shape | 16:46 |
| jrosser | did rabbitmq change their cloudsmith mirroring again? https://www.rabbitmq.com/docs/install-debian#add-a-repository-apt-source-list-file | 20:31 |
| jrosser | we could add both of those mirrors | 20:31 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!